Fraud Detecting Using Beneish M-Score and F-Score: Which is More Effective?

Beneish M-Score and F-Score are methods of identifying fraud in financial statements. The difference in research results between the two methods requires an in-depth study of the effectiveness of the Beneish M-Score and F-Score methods for detecting fraud. The type of research is descriptive and quantitative. The population in this study were companies registered in JII from 2017-2021. Through purposive sampling, ten sample companies were obtained. The results revealed that the Beneish M-Score method is more effective in detecting fraud in JII-registered companies in 2017-2021 than the F-Score. Evidenced by a higher level of accuracy and lower error type than the F-score method. However, the F-Score method has weaknesses in the level of accuracy and type error. Therefore, the Beneish M-Score helps regulators and auditors detect fraud in financial statements, so it is very important for potential investors to be able to make the right investing decisions.


INTRODUCTION
A global survey conducted by ACFE (2020) in 2020 Report To The Nation stated that asset misappropriation cases were 86% of cases with the lowest average loss of USD 100,000 per case, and asset abuse cases were the highest percentage of cases. The second highest cases were corruption, which was 43%, with the lowest average loss of USD 200,000. These corruption cases include offenses such as bribery, conflicts of interest, and extortion. In contrast, the third-highest case is financial statement fraud, which has a percentage of 10%. However, this case has a huge loss of USD 954,000. According to the Indonesian Fraud Survey (SFI), in Indonesia, there were 239 cases of fraud in 2019. The results of a survey conducted by ACFE Indonesia (2019) stated that the most adverse cases were corruption cases, as many as 167 cases (69.9%). The second-highest case was asset abuse, where this case occurred in as many as 50 cases (20.9%). While the thirdhighest is financial statement fraud cases, with 22 cases (9.2%). The large number of losses caused by fraud cases in financial statements requires serious and intense resolution so that it is expected to minimize the occurrence of fraud levels. Many things can trigger cheating, such as opportunity, rationalism, pressure, etcetera.
The existence of these various cases of fraud, stakeholders must be given identification tools that can assess the feasibility of a financial statement as a basis for decision making. There are the Beneish M-Score and the F-Score. The Beneish M-Score, according to Hugo (2019) is a method for predicting fraud in financial statements in profit management, where the ratios contained in it have been proven to have the ability to predict financial statement fraud. The higher the Beneish M-Score value in the financial statement, with a cut-off value of -2.22, the greater the likelihood that the financial statements contain fraud. Likewise, the F-Score method is a development of the Beneish M-Score model, specifically designed to allow users to get a score directly without using an index in their calculations, having a cut-off value of 1, the higher the F-Score value, the higher the risk of fraud.
Various studies were developed to detect indications of financial statement fraud. Aghghaleh (2016) showed that the F-score method is claimed to be more comprehensive and effective in detecting financial statement fraud in Malaysia. Ismawati (2019) also pointed out that the F-score method is a good model used in detecting fraud in companies listed on the IDX compared to the Beneish M-score method. Therefore, this study continues previous research to conduct empirical tests on the effectiveness of the Beneish M-Score and F-Score methods in companies listed in the Jakarta Islamic Index (JII) in 2017-2021. So the purpose of this study is to find out more effective methods of Beneish M-Score and F-Score EQUILIBRIUM, Volume 10, Number 1, 2022 in detecting fraud. This research can contribute to regulators and auditors in detecting fraud in financial statements. So potential investors need to be able to make the right decisions in investing.

Beneish M-Score Method in Detecting Financial Statement Fraud
Signal theory explains how the company should give signals to users of financial statements. This theory is the basis of the relationship of financial performance with the value of the company so that it largely determines the reputation and success of the company (Mediawati, 2018). Kusumawardhani(2014) defines fraud as an act of fraud or mistake committed by a person or organization.
Fraud indicates cheating when there is inaccuracy so that it can harm the company.
So, not only important for high returns but also must attach importance to the quality of company information. The effectiveness of the Beneish M-Score method was shown by Putri (2021) to detect financial statements from 2016-2018. Mardiharjo (2021 stated that the analysis of financial statements using the Beneish M-score is effectively used because it can show three groupings: manipulators, non-manipulators, and gray companies. Based on various studies that have been carried out, the use of the Beneish M-Score is effective in detecting financial statement fraud.

F-Score Method in Detecting Financial Statement Fraud
F-Score, according to Aghghaleh(2016) is a fraud detection method using scaled logistic probability techniques. The F-score is formulated with seven ratios, including: 1) RSST Accrual; 2) Change in receivable; 3) Change in inventory; 4) Soft assets; 5) Change in cash sales; 6) Return on assets; 7) Actual issuance of stock. Devi (2021) pointed out that pentagon fraud using the F-score method positively analyzes financial statement fraud in state-owned companies listed on the IDX in 2014-2019. Dechow (2011) revealed that the F-score method is effective for identifying and reducing misstatement activities in the future. Based on various studies that have been carried out, the use of the f-score method is effective in detecting financial statement fraud.

Comparison of Beneish M-Score and F-Score
Fraud is a severe problem in the business world. Financial statement fraud proved to be the most worrying because it involved the company's management and harmed investors. Therefore, several methods have been developed to assist regulators and auditors in detecting financial statement fraud. Beneish M-score is a statistical model using financial ratios to evaluate manipulations in financial statements. F-score develops a score from Beneish to predict which companies have material misstatements. Overall, there is a positive and significant relationship between the variables DSRI, SGI, TATA, AQI, RSST Accrual, ΔREC, ΔINV, and Soft assets to misstatements of financial statements related to fraud triangle theory, namely opportunities. Management can take advantage of the flexibility of generally accepted accounting principles to be able to carry out profit management practices, especially on accrual-based accounts. For example, increasing the value of goods sold by making earnings and false sales and increasing the value of assets by changing policies on accounting.
Meanwhile, a significant positive relationship between the GMI and LVGI variables to misstatements of financial statements is related to the fraud triangle theory, namely pressure. The increase in GMI indicates a negative signal because the company's gross profit has decreased. Pressure to realize investor expectations and expectations can encourage management to manipulate financial statements.
Meanwhile, the increase in LVGI indicates an increase in the company's burden.
Profit management is often carried out during the increase in debt to meet the requirements in the debt covenant contract.
The F-score method is claimed to be better at detecting financial statement fraud than the Beneish M-score method. The F-score method has better EQUILIBRIUM, Volume 10, Number 1, 2022 classification accuracy than the Beneish M-score method. The F-score combines variables, the discretionary accruals model, and other variables that produce a composite measure of the F-Score. The purpose of building the F-score is to develop a score that can be calculated directly from the financial statements.
The variables on the F-Score concern two things that can be seen directly in the financial statements, namely the quality of accruals proxied with RSST and financial performance proxied by RSST Accrual, ΔREC, ΔINV, ΔROA, Soft asset, ΔCASHSALES, and Issuance. (Aghghaleh, Mohamed, and Grace 2016) showed that the F-score is more comprehensive and more effective in detecting financial statement fraud in Malaysia. Ismawati (2019) shows that the F-score is a good method for detecting fraud in companies listed on the IDX compared to the Beneish M-score method. So, this research will compare the Beneish M-score's and F-score's effectiveness for detecting financial statement fraud.

RESEARCH METHOD
This research is a quantitative descriptive study. The population is an The Beneish M-Score can be calculated using the following formula: M = -4,80 + 0,920DSRI + 0,528GMI + 0,404AQI + 0,892SGI + 0,115DEPI -0,172SGAI The Beneish M-score assessment uses a cut-off value, which is a score higher than -2.22, a probability that the company applied opportunistic earnings management and misleading reporting.
While the variables contained in the F-Score consist of RSST accruals, changes in receivables, changes in inventories, percentages of soft assets, changes in cash sales, changes in return on assets, and issuance, with the formula: The result of the probability value will be divided by the unconditional probability of misstatement (probability of unconditional misstatement = 0.0037) to get the f-score value.

F-Score = Probability Value / Unconditional Probability
The F-Score uses a cut-off value of 1. if the value obtained is more than 1, it can be indicated that there is fraud in financial statements.

Beneish M-Score Method
The Beneish M-Score method is a model that has a cut-off value of -

F-Score Method
The F-Score method has a cut-off value of 1. Suppose a company has a score of less than 1. In that case, the company can be indicated as a non-manipulator company, but if the score obtained is more than 1, the company can be indicated Based on the results that have been taken into account, it can be seen that there are differences in the results of the beneish m-score and f-score. The error for the beneish m-score was found at 31.91%, or as many as 15 errors, while for the f-score, it was 34.04% or as many as 16 errors. The results confirmed the efficiency of the beneish m-score in detecting fraud by 68.09% or as many as 32 samples, while the f-score was only 65.96% or 31 samples. Thus, it can be concluded that the beneish m-score method is more suitable for detecting fraud in companies listed in the Jakarta Islamic Index in 2017-2021.

Beneish M-Score Method Is Effective in Detecting Financial Statement Fraud
Beneish m-score is a mathematical model that detects fraud in preparing company financial statements. This model involves several financial ratios to obtain a certain score to identify the possibility of financial statement fraud.
Based on research that has been carried out on ten samples of companies registered with JII in 2017-2021 with the beneish m-score using the ratios and calculations that have been presented, if the score or cut-off obtained is more than -2.22, it can be said that the company is considered a manipulator. Meanwhile, if the result or cut-off is less than -2.22, the company can be considered a nonmanipulator company. This result is consistent with research by Murdihardjo(2021), which revealed that the beneish m-score method could detect financial statement misstatements effectively by producing three groupings, namely manipulators, non-manipulators, and gray companies. Putri(2021) revealed that the beneish m-score is worthy of being used as a model to analyze financial statements of fraud in manufacturing companies listed on the IDX in 2016-2018. Septiani(2020) revealed that the beneish m-score method is accurate in detecting fraud in the financial statements of banking companies for the period 2016-2018, and the most dominant ratio in the grouping of manipulators and non-manipulators is the DSRI ratio.

F-Score Method is Effective in Detecting Financial Statement Fraud
F-Score is a financial statement fraud detection method developed using scaled logistic probability techniques. This method is a development of the beneish m-score designed so that users can find scores directly without using indices in their calculations. Based on the research of ten samples listed companies on the Jakarta Islamic Index in 2017-2021 using the f-score method with the ratios and calculations that have been presented with value criteria, if the results are less than one company is not indicated to have committed fraud/manipulation, but if the results of calculating the value are more than one, then the company is indicated to have committed fraud/manipulation. stated that the f-score method could be used effectively in detecting and reducing misstatement activity in the future. Hung(2017) also stated that the f-score method could be used in determining variables that can affect fraud in financial statements. Skousen(2009) stated that the f-score method can be used by investors as an initial screening tool in investigating fraud in financial statements.

Effective Method Used in Detecting Financial Statement Fraud
Fraud in financial statements proved to be the most worrying because it involved management and was detrimental to investors. Several methods have been developed to petrify in detecting fraud in financial statements. The two methods are beneish m-score and f-score. The beneish m-score method is said to be a manipulator if it has a score of more than -2.22, while the f-score method is considered a manipulator if it has a score of more than 1. It shows that the greater the beneish m-score or f-score of a financial statement, the greater the likelihood of fraud in financial statements.
Based on the analysis, the beneish m-score and f-score effectively detect financial statement fraud. The beneish m-score shows higher accuracy results compared to the f-score. The results showed that the efficiency of the beneish m-score in detecting fraud was 68.09%, while the f-score was only 65.96%. The result type error showed m-score is lower compared to the f-score. The type error of the beneish m-score method was 31.91%, while the f-score was 34.04%. Thus, the beneish m-score method is more suitable for detecting financial statement fraud in companies listed on the Jakarta Islamic Index in 2017-2021. However, if studied, these two methods have shortcomings, namely failing to detect fraud in the disclosure. It is because both methods have dependent on financial statement data, so if there is a qualitative material misstatement contained in the disclosure, it cannot be detected.
Based on the study's results, it is stated that this research is not in line with Ismawati's (2019) research, which reveals that the f-score method is more effectively used in detecting financial statement fraud. Aghghaleh's(2016a) research is also not in line with this study, in Aghghaleh's research states that the f-score method is more effective and more comprehensive than the m-score, while the results of this study prove that the beneish m-score is more effectively used in detecting financial statement fraud in companies listed in the Jakarta Islamic Index in 2017-2021.

CONCLUSION
Based on the results of research and data analysis, it can be concluded: The  The Beneish M-Score method is more effective than the F-Score method in detecting fraudulent financial statements on issuers registered with JII in 2017-2021. Effectiveness is stated from the higher accuracy level of the Beneish M-Score method and lower type error than the F-Score method. Therefore, the Beneish M-Score can assist regulators and auditors in detecting fraud in financial statements, so it is very important for potential investors to be able to make the right investment decisions.